**** BEGIN LOGGING AT Wed Jun 30 02:59:57 2010 Jun 30 05:34:52 mickey|zzZZzz, I added a function to libphone_utils called phone_utils_remove_filler_chars that takes a phone number and removes spaces/dashes/whatever the user chooses. I think you should add it to fsogsmd's high level functions so users of the API will be able to send "(3) 1234" as a phone number, what do you think? Jun 30 05:42:30 JaMa, Ok, got my commit access to e :P Will commit it in a moment. Please build after its done. Jun 30 05:42:52 TAsn: looking forward to e falling apart ;) Jun 30 05:43:21 TAsn: while you're at it comitting stuff to e... please fix illume2 too :P Jun 30 05:43:55 TAsn: or better still... move the shr_elm_softkey to it :-) Jun 30 05:43:56 what's wrong with illume2? devilhorns is doing a good job and if something is bad, just let him know. Jun 30 05:44:25 TAsn: yeah illume2 rocks... one thing I have though Jun 30 05:44:45 TAsn: if the quick-settings app is started (but not active) and you press power it does not get activated Jun 30 05:45:05 looks like if it would not start it all (if you forgot it already is started :P) Jun 30 05:45:46 TAsn: and yeah... right you are... shouuld have told him :U/ Jun 30 05:57:45 JaMa, ok shit, not there yet, will do it when I'm back. Jun 30 05:57:50 mrmoku|kiga, will read it when I'm back, ciao. Jun 30 06:26:48 dent: I confirmed that the speaker thingie is not an exclusive issue of yourself (and forgot to tell you then :/) Jun 30 06:27:05 I have it too... and mute doesn't work either... it just does nothing Jun 30 06:33:39 TAsn: good, I wasn't that quick today (after 3hours of sleep :)) Jun 30 08:45:25 _o/ Jun 30 08:47:12 gm ThibG Jun 30 08:48:01 ok, guys Jun 30 08:48:15 what do you think of implementing a plugin for ewm Jun 30 08:48:26 which allows 3d window switching? Jun 30 08:48:28 :-) Jun 30 08:48:56 means, that we have 3d app selection in SHR on platforms which support it Jun 30 08:49:03 (dri/gl) Jun 30 08:49:37 hehe first real-world illume2-launcher replacement? :) Jun 30 08:50:08 something like beryl/compiz for smartphones Jun 30 08:50:26 apple has it, bada has it Jun 30 08:50:39 my laptop and workstation have it Jun 30 08:50:52 why not the phone too? ;-) Jun 30 08:51:14 http://en.wikipedia.org/wiki/Bada_%28operating_system%29 Jun 30 08:51:19 bada linux btw Jun 30 08:51:22 if it works as fast as illume2 launcher, then why not :P Jun 30 08:52:00 perhaps we could mofify illume2 to do it there? Jun 30 08:52:24 we could then make a configuration setting for disabling/allowing it Jun 30 08:52:34 probably yes as main difference between illume1 and illume2 is that later are replacable easier Jun 30 08:53:21 raster even said that those illume2 parts included are just "demos" Jun 30 08:53:45 but good enough for SHR now :) Jun 30 08:55:53 hmm Jun 30 08:56:09 we should actually start to extend the graphics part too Jun 30 08:56:11 :-) Jun 30 08:56:30 as I've heard, we have elementary now Jun 30 08:56:43 but nearly none of the apps we have is using them Jun 30 08:56:47 * JaMa just extended xserver 1.9 to work on nonKMS DDX again :) Jun 30 08:57:03 aha Jun 30 08:57:05 ok Jun 30 08:57:11 do we have rotation with fb? Jun 30 08:57:14 on xorg Jun 30 08:57:31 I mean xrandr support Jun 30 08:57:37 (x) Jun 30 08:57:59 you know what I mean Jun 30 08:58:01 -_- Jun 30 08:58:03 JaMa: :D Jun 30 08:58:26 leviathan: maemo has it :) Jun 30 08:58:38 Weiss: pNew->name = NULL wasn't enough, s/xnfalloc(/xnfcalloc(1,/ seems to do the job right :) Jun 30 08:58:41 leviathan: basically, you're talking about making Illume into a compositing window manager with 3D effects Jun 30 08:58:55 Weiss: and glamo-kms-output was already using calloc, but old glamo-output didn't Jun 30 08:59:18 JaMa: aha, ok.. so something wasn't zeroed which should have been? Jun 30 08:59:34 Weiss: e already has that Jun 30 08:59:35 :) Jun 30 09:00:08 Weiss: yes Jun 30 09:00:14 I do exactly that Jun 30 09:00:16 :-) Jun 30 09:00:16 Weiss: yes and that failing free() was just one part of it ("No valid mode" was also caused by that) Jun 30 09:01:12 Weiss: what do you think about the idea to extend illumes capabilities to composite window managment with 3d effects? Jun 30 09:02:51 raster: compositing? Jun 30 09:04:25 as I know they already implemented it on BADA under enlightenment, so perhaps you wont get an answer, because he is not allowed because of the NDA Jun 30 09:04:29 ^^ Jun 30 09:04:37 Weiss: tried the comp module? Jun 30 09:04:46 hmm Jun 30 09:04:51 comp module? Jun 30 09:05:01 leviathan: HAHAHHAHAHA and who told u that? Jun 30 09:05:15 just an idea Jun 30 09:05:20 leviathan: see e17. there is a comp (compositor) module Jun 30 09:05:25 because I've seen 3d on the new galaxy Jun 30 09:05:28 it works wityh software compositing ore opengl Jun 30 09:05:31 or wave or how this is called Jun 30 09:05:37 opengl;-es2 works - with texture-from-pixmap Jun 30 09:05:41 works on desktop gl too Jun 30 09:05:42 ok Jun 30 09:05:46 evas gl enigne Jun 30 09:05:53 I'll try it Jun 30 09:05:57 where can I find it? Jun 30 09:06:01 leviathan: galaxy is android Jun 30 09:06:05 wave is bada Jun 30 09:06:13 bada is a middleware layer on top of a nucleus rtos Jun 30 09:06:25 raster: cool :D Jun 30 09:06:27 i know nothing of that sutff that isnt already public Jun 30 09:06:32 so nothing nda to talk about Jun 30 09:06:38 all the e stuff is all in upsteram Jun 30 09:06:43 inm svn - in out latest snapshots Jun 30 09:06:44 all there. Jun 30 09:06:46 I will suggest soon, to replace the middleware by FSO Jun 30 09:06:51 to save development costs Jun 30 09:07:12 e has nothing to do with bada in any way or form Jun 30 09:07:46 sry. I'm not developing this x't next generation linux for smartphones Jun 30 09:07:53 they shall use SHR+FSO Jun 30 09:07:54 -_- Jun 30 09:12:17 okeee Jun 30 09:12:18 nice Jun 30 09:12:44 veery nice Jun 30 09:12:55 imagine that in combination with moko-suite Jun 30 09:13:01 O_o Jun 30 09:13:14 larsc, ping Jun 30 09:14:11 raster: good job Jun 30 09:14:56 what we woul need now is something exposé like Jun 30 09:22:20 ok Jun 30 09:22:34 I've got enoug Jun 30 09:22:48 I'll wait for daniel walker to finish the graphics Jun 30 09:22:51 -.- Jun 30 09:23:00 ok Jun 30 09:23:06 misc: w701? Jun 30 09:23:11 what device is that? Jun 30 09:23:13 * mrmoku was wondering too Jun 30 09:23:13 new laptop monste Jun 30 09:23:14 r Jun 30 09:23:19 ahh :P Jun 30 09:23:23 ahh Jun 30 09:23:24 ok Jun 30 09:23:25 17" beast Jun 30 09:23:26 ^^ Jun 30 09:23:30 almost a workstation Jun 30 09:23:51 I've got two lcd widescreens in front of me Jun 30 09:23:54 really nice Jun 30 09:23:58 :-) Jun 30 09:24:05 btw Jun 30 09:24:21 the tx of the htcdream seems gone Jun 30 09:24:26 -.- Jun 30 09:24:28 d'oh, broken? Jun 30 09:24:34 I fear so Jun 30 09:24:39 I'm so ennoyed Jun 30 09:24:44 I'll wait for walker Jun 30 09:24:46 ... Jun 30 09:24:53 i can imagine. what did you wanted to use the uart for? Jun 30 09:25:13 to not be forced to wate for walker Jun 30 09:25:28 until he has graphics working Jun 30 09:25:51 thats something important I'll have to say to the samsung developers Jun 30 09:26:04 this chip, and this... ah and zener diodes to the uart Jun 30 09:26:06 -_- Jun 30 09:27:08 the ft232 is still functional Jun 30 09:27:16 I tried by connecting rx with tx Jun 30 09:27:27 but the transmit channel of the phone is gone Jun 30 09:27:35 I can only recieve data Jun 30 09:27:45 thats a piece of shit Jun 30 09:27:50 this device at all Jun 30 09:28:11 android unusable, and the hardware design is flaw Jun 30 09:28:40 at least I know now, where all these sleeping students in the lecture room will get a job later... -.- Jun 30 09:29:03 I'll go on try'n to get a better phone, in cooperation with samsung Jun 30 09:29:14 mickey|w701: btw, whats your address? Jun 30 09:29:17 :-) Jun 30 09:29:32 * leviathan needs it for sending development phone Jun 30 09:29:32 shipping? see vanille-media.de -> impressum Jun 30 09:29:36 ok Jun 30 09:30:23 heh Jun 30 09:30:29 * mrmoku got WS on boot now :/ Jun 30 09:31:29 mrmoku: let it sit for a minute or two, that 'fixes' it for me Jun 30 09:32:00 blindcoder: ok... fortunatelly I need only ssh for now Jun 30 09:32:36 mickey|w701: btw, I found out now, that we have already 3d in SHR Jun 30 09:32:39 ^^ Jun 30 09:33:08 on which machine` Jun 30 09:33:10 ? Jun 30 09:33:50 sw rendering Jun 30 09:33:54 so everyone Jun 30 09:34:03 I ment more in respect to composite Jun 30 09:34:05 :-) Jun 30 09:34:09 you know Jun 30 09:34:11 compiz Jun 30 09:34:21 we have such thing in SHR Jun 30 09:34:24 very nice Jun 30 09:34:29 SW rendering on s3c2442 is horribly slow because of no FPU, though Jun 30 09:34:30 now we need exposé Jun 30 09:34:38 ^^ Jun 30 09:34:53 as I remember playya showed me something Jun 30 09:34:57 called gadget Jun 30 09:34:59 which do this Jun 30 09:35:10 but: proper compositing allows X to make good use of acceleration - it's what EXA was designed for Jun 30 09:35:26 render a window once, then fling it round the screen - should be fast Jun 30 09:36:18 GNUtoo|laptop: tx is gone Jun 30 09:36:21 on the phone Jun 30 09:36:29 leviathan, ok Jun 30 09:36:30 I should have 2.8 V on each side Jun 30 09:36:34 tx and rx Jun 30 09:36:39 bbias, need to reboot Jun 30 09:36:39 but tx is 0V Jun 30 09:36:41 :( Jun 30 09:36:50 so its broken Jun 30 09:37:00 I'll eat Jun 30 09:37:03 brb Jun 30 09:37:04 what connector did you use? Jun 30 09:37:12 uhm Jun 30 09:37:21 I checked for short circuits Jun 30 09:37:22 if it's an audio one Jun 30 09:37:37 I fear, its because I connected it the wrong way around first Jun 30 09:37:41 the tx and ground connector are connected Jun 30 09:37:43 a short try was all it needed Jun 30 09:37:59 hmm? Jun 30 09:38:19 but I have 3.3V on tx and rx on usb side Jun 30 09:38:23 mom Jun 30 09:38:26 I'll cook Jun 30 09:38:31 I'll come back with the food Jun 30 09:38:33 10min? Jun 30 09:38:35 ok Jun 30 09:49:26 hello! anyone having corruption of illume home screen after suspend/resume with recent shr-unstable image? Jun 30 09:55:14 mrmoku, is your screen still white? Jun 30 10:03:11 ok Jun 30 10:03:14 took bit longer Jun 30 10:03:20 but its very delicious Jun 30 10:03:23 GNUtoo|laptop: re Jun 30 10:03:24 ^^ Jun 30 10:03:42 leviathan, ok Jun 30 10:03:44 tx seems broken Jun 30 10:03:52 try the connector without the dream connected Jun 30 10:04:08 then I have 3.3V Jun 30 10:04:18 with the dream connected I've got 0V Jun 30 10:05:18 it seems that tx does now a short circuiting to mass Jun 30 10:05:21 internally Jun 30 10:05:23 what means Jun 30 10:05:26 uart is broken Jun 30 10:06:36 what means spit on qualcomm i wanna have samsung on board Jun 30 10:06:38 -_- Jun 30 10:14:14 ThibG: white as snow :P Jun 30 10:14:28 ThibG: though that is on om-gta02-2.6.32 without any patches Jun 30 10:14:33 might just be normal :) Jun 30 10:15:49 ThibG: I'm trying to fix that jack insert input thing... http://pastie.org/1024803 is not enough :P Jun 30 10:16:02 mrmoku, can you cat /sys/whatever/glamoSOMENUMBERS/regs? Jun 30 10:16:36 ThibG: root@om-gta02 ~ # cat /sys/devices/platform/s3c2440-i2c/i2c-0/0-0073/glamo3362.0/regs Jun 30 10:16:39 Segmentation fault Jun 30 10:16:47 oh, yes, forgot about that :x Jun 30 10:32:04 Weiss: please apply when you have time&keys http://gitorious.org/~jama/angstrom/jama-shr-experimental/blobs/org.openembedded.dev/recipes/xorg-driver/xf86-video-glamo/0001-glamo-output-initialize-mode-name-to-NULL.patch Jun 30 11:04:55 Weiss: i wonder if with fifo patch we can use dma efficltively. Jun 30 11:07:24 Weiss: feed it with n words, then wait, feed next bulk. while wait other processes can use memory bus. Jun 30 11:20:38 Weiss: this way decoder may use almost all cpu, as currently profile of deconing is like this: in xorg, 96% in mempy, 2% in kernel. 7% in glamo drv. in mplayer: 49 in awcodec, 29 in swscale(seem colorspace conversions), 4.7 kernel, 4.6 dl loading(n/a). for overall 48% idle/29% mplayer/18%org. Jun 30 11:21:46 Weiss: this way decoder may use almost all cpu, as currently profile of decoding is like this: in xorg, 96% in memcpy, 2% in kernel. 0.49% in glamo drv. in mplayer: 49 in awcodec, 29 in swscale(seem colorspace conversions), 4.7 kernel, 4.6 dl loading(n/a). for overall 48% idle/29% mplayer/18%org. Jun 30 11:24:25 gena2x: doubt it... the problem isn't so much the speed, but the locking of the bus. no matter what you do, the memory bus ends up locked Jun 30 11:25:30 Weiss: even with fifo? Jun 30 11:26:36 Weiss: oh, or fifo between glamo or display... Jun 30 11:26:40 yep.. the FIFO is between the LCM and VRAM, and the memory bus is between VRAM and SoC Jun 30 11:29:18 Weiss: interesting, why glamo bus is so strange? to be 7mb/s it should have 1.8Mb/s of words. why it is connected to memory bus in this case? may be it has some kind of fifo too? Jun 30 11:30:45 gena2x: you can find the full story (in bits) in old mailing list posts, but basically Glamo abuses a flag on the memory bus which is meant only for arbitration failures. it uses it whenever its own FIFO (different FIFO) fills up because it can't process input fast enough Jun 30 11:32:16 Weiss: sounds very good. can we do fast transfer until this internal fifo full, then wait, the do next transfer? can you recall and point me to thread to read? Jun 30 11:33:20 Weiss: dma controller seem can do all this for us. Jun 30 11:36:31 http://www.mail-archive.com/openmoko-kernel@lists.openmoko.org/msg03657.html - read the whole thread - I think it covers most of your ideas.. Jun 30 11:36:46 Weiss: thank you! Jun 30 11:38:02 gena2x: maybe bandwidth to Glamo increased with the LCM FIFO patch.. might be worth testing Jun 30 11:38:20 but I read that we're already in excess of what SMedia think should be possible Jun 30 11:41:40 Weiss: if glamo was some fifo on memory bus, and nobody cares about this, we can do amaizing things... Jun 30 11:43:19 hmm? Jun 30 11:44:02 Weiss: no need to go beyond speed limit, idea is to full it's fifo, wait, full again. as i guess that while fifo is not full we can transfer data to glamo at bus speed? Jun 30 11:44:50 I think so... but I think that FIFO (if it even exists) is small.. maybe only 16 bits Jun 30 11:45:50 Weiss: hm... i thought memory bus is 4 byte!? how it is possible to have 16 bit fifo? Jun 30 11:46:05 Weiss: or you mean 16 bytes? Jun 30 11:47:28 Weiss: as i told, dma controller can solve this problem for us. it can idependently transfer data in 16 byte bulks, then wait. Jun 30 11:47:38 gena2x: transfers are 16 bits according to those mailing list posts.. I'd have to check the schematics to be sure Jun 30 11:47:58 Weiss: oh, i didn't read that all. Jun 30 11:47:58 but, the DMA controller is on the same bus.. there's no separate "shipping lane" Jun 30 11:48:10 Weiss: this is not a problem. Jun 30 11:48:16 Weiss: as we have fifo. Jun 30 11:49:17 who knows exactly what goes on inside Glamo.. maybe there is no FIFO. there certainly doesn't seem to be a usefully large one.. Jun 30 11:49:33 "The low parts of the attached picture is when the Glamo has stalled the Jun 30 11:49:39 CPU from the external bus (shared with DRAM) during the scrolling Jun 30 11:49:39 action, the high parts are when it was willing to take another 16-bit Jun 30 11:49:40 transfer Jun 30 11:49:55 Weiss: ok, i'll try this. thanks for links and description! Jun 30 11:50:19 no problem Jun 30 11:50:34 would be nice to find more speedups.. but I think the previous work is pretty comprehensive here.. Jun 30 11:51:28 Weiss: your finding about FIFO proves that it's possible to do much more. as did my nodebug thing. Jun 30 11:52:15 Weiss: i think originally om people has too much work and issues, that they easily may overlook something. Jun 30 11:53:29 Weiss: and i still believe glamo developers are really smart. Jun 30 11:54:16 Weiss: so if their device can do only 7mb/s it is too stupid to connect device to memory bus. Jun 30 11:56:13 which is 133 MB/s... Jun 30 11:57:48 maybe.. Jun 30 11:58:08 oh, why we can't get comments from them? Jun 30 11:58:14 from SMedia? Jun 30 11:58:21 om did to glamo so much bad PR... Jun 30 11:58:36 from people who delevoped glamo itself. Jun 30 11:58:50 not from marketing/director. Jun 30 12:00:17 ok, gone to read thread.... Jun 30 12:00:50 SMedia (the company) got eaten up by another company AFAIK Jun 30 12:01:09 but people not dead, yes? :) Jun 30 12:02:08 who knows Jun 30 12:02:18 :) Jun 30 12:03:02 i hope thier country has no tradition to fire in derect meaning of word... Jun 30 12:08:43 Hey Jenkins.. You're fired *pulls out a match*. Jun 30 12:28:24 ThibG: pong Jun 30 12:30:03 in glamo-mci, slow_ratio and glamo_mmc_use_slow are not used anymore Jun 30 12:30:06 is that intended? Jun 30 12:34:10 well, more or less. Jun 30 12:37:29 hm? Jun 30 12:39:49 i think it was hooked up to a function telling whether the gps is enabled or not Jun 30 12:41:25 yes Jun 30 12:41:37 the function is still registred in mach-gta02.c Jun 30 12:41:49 but it doesn't get called anymore Jun 30 12:43:53 larsc: hi, what about gta02-resume reason, it was available in 2.6.29 and gone in 2.6.3x kernels, is it replaced with something? Jun 30 12:44:17 JaMa|Off: i have no idea Jun 30 12:48:06 larsc: it was in arch/arm/mach-s3c2442/mach-gta02.c as platform_device gta02_resume_reason_device Jun 30 12:48:35 JaMa|Off: ahh... that explains why I don't see that in fsodeviced.log anymore :P Jun 30 12:49:12 JaMa|Off: ./devices/platform/s3c2440-i2c/i2c-0/0-0073/resume_reason Jun 30 12:50:18 yup there is only subreason (as fsousaged call it) Jun 30 12:51:22 and subreason is read only if ( reasonvalue == ResumeReason.PMU ) Jun 30 12:58:41 the resume reason driver looks like a hack Jun 30 13:01:45 ha.. Jun 30 13:03:48 JaMa|Off: what is it used for? Jun 30 13:07:13 i'm not even sure that it reports any meaningfull value Jun 30 13:07:23 would report Jun 30 13:09:33 larsc: fsousaged reports why the phone was resumed, I guess it worked before http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsousaged/src/plugins/lowlevel_openmoko/plugin.vala;h=91152d6df65ec243c101f2822f3c2958ef436c04;hb=f426a51eced649e27cefa43403c3ab90b4573ebc Jun 30 13:14:08 but is this info used for anything meaningfull? Jun 30 13:14:34 not present, since UI is lacking Jun 30 13:15:10 but that's not a reason not to export it from the middleware Jun 30 13:18:08 mickey|w701: Q-Master said yesterday something about the need to let calypso know the reason, is that right? Jun 30 13:18:48 JaMa|Off: no, we don't do anything like that Jun 30 13:19:04 different question: could it be used for anything useful? Jun 30 13:19:23 sure Jun 30 13:19:46 if power button is the wakeup reason we turn on the screen immediately Jun 30 13:19:47 mrmoku: good Jun 30 13:19:51 if gsm is the wakeup reason, we don't Jun 30 13:19:51 mickey|w701: good :) Jun 30 13:20:18 but rather check whether the incoming SMS is from someone important Jun 30 13:20:23 and if not, fall asleep again Jun 30 13:24:37 Weiss: ~ - the Glamo paces US according to its own private arbitration of the Jun 30 13:24:37 internal DRAM between us and, eg, the video controller inside Glamo that Jun 30 13:24:37 is also constantly needing access to same RAM to update the LCM Jun 30 13:24:37 " - not true animore after your patch. Jun 30 13:25:05 gena2x: well.. certainly a little less true Jun 30 13:27:03 "~ - we pace ourselves at the CPU with waitstates to be REALLY slow if we Jun 30 13:27:03 touch Glamo. If we don't do that we get unstable communication and Jun 30 13:27:03 visual artefacts. Jun 30 13:27:03 " sounds like just exprirement "we tried, it fails, we can't set it :(". Jun 30 13:28:41 no real attempt to understand why. (i didn't finish reading...) Jun 30 13:40:18 "degraded more as the dma block sizes went down." may be caused by memory bulk read performance gains. "SOC were blocked until DMA was done" or course!!! whooh. some misteries! Jun 30 13:40:43 and then they speak exactly about my idea... Jun 30 13:43:47 but nobody tried it. Jun 30 13:43:56 * gena2x finished. Jun 30 13:44:25 so, nobody actually tried this. Jun 30 13:44:38 they used DMA as memcpy. Jun 30 13:44:59 lets try :) Jun 30 13:46:02 * gena2x always wanted to test dma operation :) Jun 30 13:47:00 what do you want to do? Jun 30 13:48:01 larsc: dma tranfer to glamo in small bulks. Jun 30 13:49:11 larsc: i also want to measure dma memcpy speed. Jun 30 13:49:53 larsc: and check if it may be useful to use dma for large memcpys. Jun 30 13:50:42 larsc: after profiling, i found that our cpy is almost always busy with memcpy. Jun 30 13:50:59 larsc: in Xorg (to glamo), in usual apps (mem2mem) Jun 30 13:51:10 in Xorg, it is 98%. Jun 30 13:52:39 larsc: how do think, is it nice idea to check? Jun 30 13:52:47 larsc: how do you think, is it nice idea to check? Jun 30 13:53:49 sure Jun 30 13:58:17 I'm sure I remember more discussion.. think it was in a Trac ticket somewhere, about someone actually trying "DMA on timer", though Jun 30 13:59:00 Weiss: lets recheck this, anyway, now with your patch. Jun 30 13:59:40 and without debuggings interrupts... Jun 30 14:00:07 yeah, good idea Jun 30 14:19:44 GNUtoo|laptop: today I came to a conclusion Jun 30 14:19:50 I hate android AND qualcomm -_- Jun 30 14:20:03 this fucking txd is broken Jun 30 14:20:16 I can't even recieve my own symbols Jun 30 14:20:26 when short cutting rx and tx Jun 30 14:20:31 its kind of crap Jun 30 14:20:54 theres exactly one 2.8V pin Jun 30 14:20:58 the rest is all 0V Jun 30 14:20:59 -_- Jun 30 14:21:28 so either I use jtag Jun 30 14:21:35 or I repair the ttls on the board Jun 30 14:21:42 both options are more then difficult Jun 30 14:22:12 I guess I'll need to organize this meeting very soon Jun 30 14:22:26 hmm, OMAP3 is supported by OE Jun 30 14:22:38 mickey|w701: do you have a samsung phone? Jun 30 14:22:53 and if yes, which one? Jun 30 14:23:01 leviathan, yes omap3 is supported very well in oe Jun 30 14:23:06 good Jun 30 14:23:13 because samsung is using OMAP3 Jun 30 14:23:19 yes Jun 30 14:24:08 I need a phone, where I can review the implementation of the drivers Jun 30 14:24:12 else I can't trust it Jun 30 14:24:16 and so I can't use it Jun 30 14:24:39 qualcomm has already shown its pleasentness about my try's to make theire driver usable Jun 30 14:24:42 so... Jun 30 14:24:44 -.- Jun 30 14:24:51 I'll try it with the concurrence Jun 30 14:24:53 :-) Jun 30 14:25:34 SHR seems nice Jun 30 14:25:44 also for business men Jun 30 14:25:49 with the composite enabled Jun 30 14:25:56 you should try ^_^ Jun 30 14:26:14 so we have already effects in SHR Jun 30 14:26:33 we have exposé in the gadgets part of the configuration too Jun 30 14:26:41 in shr2 (illume2) Jun 30 14:27:48 leviathan, try palm pre,it's omap and will soon be suported,ask mickeyl and morphis Jun 30 14:27:49 mickey|w701: do you have a samsung smartphone? Jun 30 14:27:58 hmm, ok Jun 30 14:28:04 palm pre Jun 30 14:28:10 or palm pre plus Jun 30 14:28:17 but palm pre is safer Jun 30 14:29:30 # CHF 649.- Jun 30 14:29:37 woaaa Jun 30 14:29:56 thats expensiver then my netbook -_- Jun 30 14:31:07 uhm Jun 30 14:31:09 oops Jun 30 14:31:17 expensiver doesnt exist in english Jun 30 14:31:22 I ment more expensiv Jun 30 14:31:23 sry Jun 30 14:31:25 hmm Jun 30 14:32:17 GNUtoo|laptop: do you have a palm pre? Jun 30 14:32:30 no Jun 30 14:32:54 how do you help developing there? Jun 30 14:32:56 I'm not shure if I'll buy one Jun 30 14:33:01 hmm Jun 30 14:33:09 If I fix the htcdream Jun 30 14:33:16 I'll hesitate even more Jun 30 14:33:27 btw, GNUtoo|laptop for shipping a development samsung phone to you Jun 30 14:33:34 they will need your address Jun 30 14:33:35 ^^ Jun 30 14:33:45 ok, we'll talk about that later Jun 30 14:33:49 jup Jun 30 14:33:58 first I need something to show Jun 30 14:34:43 perhaps I will bring freerunner and htcdream, and tell them the lack of functionality is only because of the missing information about the hardware Jun 30 14:35:01 and they could spare a lot of money, if they let us do the development Jun 30 14:35:17 because we would just add fso support Jun 30 14:35:30 and compile an image for omap3 Jun 30 14:35:46 then they don't have to deploy a developers crew Jun 30 14:35:49 for each new phone Jun 30 14:36:04 but can only sponser new phones to us regulary ^_^ Jun 30 14:36:50 okee, we need a concept so that we can make a deal Jun 30 14:36:58 where everyone is willing to agree Jun 30 14:37:10 my company makes support Jun 30 14:37:16 you're getting phones sponsored Jun 30 14:37:34 hi all! Jun 30 14:37:37 hi Jun 30 14:37:48 JaMa|Off: mrmoku: any progress on headset? Jun 30 14:38:05 also after suspend i may have issues with render artifacts with 2.6.32 Jun 30 14:38:18 and unavailable sdhc card Jun 30 14:43:29 heyho Jun 30 14:43:49 mrmoku, i just send a patch to cleanup the keypad widget to the shr-devel ML Jun 30 14:44:06 mrmoku, if you have some free time, could you please review and apply? Jun 30 14:47:44 hmm seems like i'm not allowed to post to shr-devel.. Jun 30 14:47:49 Slyon: great, will try to take a look this evening Jun 30 14:47:50 oh Jun 30 14:48:41 alexxy: I tried to (re-)add the input device for it... not yet fully functional Jun 30 14:48:44 some bit is missing :/ Jun 30 14:48:55 ok Jun 30 14:49:44 ThibG: http://pastie.org/1024803 gives me the input device and I can see the interrupts for jack in /proc/interrupts... but no input event Jun 30 14:49:52 ThibG: any idea which piece might be missing? Jun 30 14:49:56 GNUtoo|laptop: I've got the idea Jun 30 14:50:04 I'll show them SHR on the beamer Jun 30 14:50:08 in an emulator Jun 30 14:50:14 ??? Jun 30 14:50:23 I'll emulate OMAP3 Jun 30 14:50:27 really? Jun 30 14:50:28 so we can be sure, that its working Jun 30 14:50:32 and also Jun 30 14:50:45 you are about to throw a lot of money on the thing Jun 30 14:50:58 would they like freerunner, palm or what ever? Jun 30 14:51:00 and you say that the pre is too expensive? Jun 30 14:51:03 first they need to buy it Jun 30 14:51:14 else, we can't throw a lot of money into it ;-) Jun 30 14:54:47 mrmoku, hm Jun 30 14:54:52 * mrmoku tries .code = SW_HEADPHONE_INSERT Jun 30 14:55:09 that was done in neo1973kbd.c Jun 30 14:55:25 not sure how it is done now Jun 30 14:55:27 yeah, and is standard gpio-keys now (the Aux part) Jun 30 14:56:16 just adding it to the gpio-buttons gives me the additional input 'key' (seen in evtest) and the interrupts Jun 30 14:58:26 ThibG: hehe, SW_HEADPHONE_INSERT did the trick :D Jun 30 14:58:42 :) Jun 30 14:58:51 [SIGNAL] org.freesmartphone.Device.Input.Event /org/freesmartphone/Device/Input :1.2 Jun 30 14:58:54 ( "HEADSET", "released", 25 ) Jun 30 14:58:57 :D Jun 30 14:59:07 mrmoku, any idea what could be wrong if i'm suscribed to the list but not allowed to post? Jun 30 14:59:28 Slyon: patch too big could be one reason? Jun 30 15:00:38 mrmoku, is 32kb too big? Jun 30 15:05:49 Slyon: hmm... don't remember where the limit is Jun 30 15:08:17 mrmoku, anyways.. i have to run. just pastet the patch, as git send-email doesn't work: http://shr.pastebin.com/JbGzWuiE Jun 30 15:08:29 bye! Jun 30 15:09:05 Slyon: ok Jun 30 15:09:23 GNUtoo|laptop: http://code.google.com/p/qemu-omap3/wiki/UserManual Jun 30 15:09:36 GNUtoo|laptop: its not my issue, but the one of my company Jun 30 15:09:46 ok Jun 30 15:09:52 I'll need to declare, why I need money for an ADDITIONAL smartphone Jun 30 15:09:58 ok Jun 30 15:10:01 while I've got already 3 functional ones Jun 30 15:10:05 at least Jun 30 15:10:07 lol ok Jun 30 15:10:07 no more Jun 30 15:10:10 =) Jun 30 15:10:15 2 freerunner Jun 30 15:10:19 1 htcdream Jun 30 15:10:23 1 sony ericson Jun 30 15:10:25 1 nokia Jun 30 15:10:27 1 n810 Jun 30 15:10:30 1 laptop Jun 30 15:10:37 1 6-core pc Jun 30 15:10:40 etc. Jun 30 15:10:43 ok Jun 30 15:11:05 oh, I forgot the c64 and jornada like handhelds Jun 30 15:11:07 hmmm Jun 30 15:11:12 I'll do it with this Jun 30 15:11:20 http://code.google.com/p/qemu-omap3/wiki/UserManual Jun 30 15:11:36 then everyone can see it too Jun 30 15:11:38 :-) Jun 30 15:11:50 on a beamer Jun 30 15:12:04 then I show them, that I'm already using it on my personal phone Jun 30 15:12:17 but because we do not have access to the whole documentation Jun 30 15:12:22 its not fully functional Jun 30 15:12:41 and we would like to make it run with full capability trough theire cooperation Jun 30 15:12:53 so that every one can take profit Jun 30 15:13:03 we have a fully supported SHR phone Jun 30 15:13:10 they spare development costs and time Jun 30 15:13:16 everyone happy Jun 30 15:13:20 and I gotte go Jun 30 15:13:27 computer club meeting Jun 30 15:13:28 ^^ Jun 30 15:29:36 hi, Jun 30 15:31:22 is there some kernel hacker here, I've that code: http://pastebin.com/HWHDv9q9 which crash in the alsa_send_silent_buffer function at wait_event_interruptible Jun 30 15:31:32 I was told to implement workqueues instead Jun 30 15:31:46 but the good implementation doesn't come Jun 30 15:31:53 what should I do? Jun 30 15:32:19 the crash is this one: Jun 30 15:32:27 [ 63.489013] BUG: scheduling while atomic: mplayer/885/0x00000003 Jun 30 15:36:19 ThibG: sent a patch to the kernel list... who can commit to om-gta02-2.6.32 branch? Jun 30 15:36:36 larsc, at least Jun 30 15:37:13 larsc: do you still commit to 2.6.32 too? :P Jun 30 15:37:42 JaMa|Off: otherwise we could just add it to our list of patches... Jun 30 15:39:32 PaulFertser: ping Jun 30 15:41:10 mrmoku: read your mails before trying to merge your jack patch. anarsoul is right. if 2.6.32 has soc-jack, you've better use it imho :) Jun 30 15:42:25 rtp: already reading in Documentation/sound/soc/jack.txt :) Jun 30 15:43:00 ok :) Jun 30 15:45:40 mrmoku: nak Jun 30 15:46:14 btw. it's already exported through the sound-jack api... Jun 30 15:48:45 or it least it did at some point... Jun 30 15:48:48 hm Jun 30 15:51:06 mrmoku: https://dev.openwrt.org/changeset/17069#file1 Jun 30 15:51:59 larsc: ic... where does one see the jack events in userspace? Jun 30 15:52:30 /dev/input/eventX Jun 30 15:52:57 for whatever reason that patch isn't in the openmoko kernel tree though Jun 30 15:52:57 larsc: then something is missing... we have 4 input devices and none reports jack events Jun 30 15:53:00 ahh Jun 30 15:53:02 ok :) Jun 30 15:53:12 anyone? how do I debug that or implement something better? Jun 30 16:37:50 JaMa|Off, sec, not there yet :( committing now. Jun 30 16:51:58 FINALLY!!! Jun 30 16:52:03 JaMa|Off, I just committed my patch :P Jun 30 16:54:04 JaMa|Off, 49949 have fun :P Jun 30 16:55:26 TAsn: do we have new enough fribidi? Jun 30 16:55:45 TAsn: shouldn't it be checked in configure.ac (maybe it is) Jun 30 16:55:51 let me verify. Jun 30 16:56:13 as for the check: it doesn't depend on a new version of fribidi Jun 30 16:56:23 the new version just automatically does (with the new API) Jun 30 16:56:26 things the old one didn't. Jun 30 16:56:33 so no need to check for version in configure.ac Jun 30 16:56:35 hmm no version requirement Jun 30 16:56:37 and you can build Jun 30 16:56:50 in the future Jun 30 16:56:56 I do intend to use the new API Jun 30 16:57:00 ok then Jun 30 16:57:09 and will request a specific version Jun 30 16:59:42 * JaMa|Off building Jun 30 17:02:09 TAsn: looks like debug always on :/ http://trac.enlightenment.org/e/changeset/49947 Jun 30 17:11:06 JaMa|Off, only in elmentary's cut and paste Jun 30 17:11:15 and I bet it's because they want useful debug info Jun 30 17:11:18 as it does crash sometime Jun 30 17:11:57 JaMa|Off, So please build, don't look for excuses!!! :P Jun 30 17:18:01 local build is running :) NOTE: Running task 2141 of 11953 Jun 30 17:18:24 :( Jun 30 17:18:49 * JaMa|Off changing scaling_governor to performance and slowly thinking about leaving work Jun 30 17:25:16 JaMa|Off, Compiling on your laptop?! Jun 30 17:26:17 TAsn: no.. desktop Jun 30 17:28:48 so why are you scaling your cpu? Jun 30 17:29:15 power consumption when I don't need power Jun 30 17:29:43 it's quite a difference 4cores on 3400Mhz or 800Mhz.. Jun 30 17:30:06 AHH :P Jun 30 17:30:07 ok, leaving now.. cya Jun 30 17:30:10 4 cores :P Jun 30 17:30:26 Yay, reflashed and now it works! Jun 30 17:30:27 ciao. Jun 30 17:30:29 (gsm) Jun 30 17:30:36 TAsn: if you have time... could you take a look at Slyons patch? Jun 30 17:30:53 of course Jun 30 17:30:55 where is it at? Jun 30 17:32:53 mrmoku, cleanup keypad? Jun 30 17:33:40 ok applying Jun 30 17:33:44 his patch is very good. Jun 30 17:34:09 mrmoku, but please do me a favor, and clean up patchwork :P I don't have access. Jun 30 17:34:25 Already applied all of david's patches Jun 30 17:34:29 :P Jun 30 17:34:40 And mark slyon's as applied as well please. Jun 30 17:37:20 SHR: 03lukasmaerdian 07libphone-ui-shr * rccbacb06caec 10/data/keypad.edc: cleanup keypad widget Jun 30 17:37:39 JaMa|Off, no special need to rebuild though Jun 30 17:37:46 as it's a cosmetic patch Jun 30 17:45:35 TAsn: ok Jun 30 17:46:58 TAsn: all 4? or have the first two been reworked? Jun 30 17:47:44 nvm Jun 30 17:49:50 the first two are bad :P Jun 30 17:58:49 mrmoku, btw, did you see the index2 example in elm? I wonder what it does :P Jun 30 18:41:45 gnutoo: heyo Jun 30 18:41:57 hi Jun 30 18:42:46 I'll build an beagleboard SHR image Jun 30 18:43:49 leviathan, :)) Jun 30 18:43:52 SHR: 03tom 07libphone-utils * rd3f510d75f76 10/ (4 files in 4 dirs): phoneutils_test now compiles with the rest of the project. Jun 30 18:44:43 TAsn: no, did not see index2 Jun 30 18:44:44 TAsn: hi Jun 30 18:45:07 mrmoku, should be cool :P Jun 30 18:45:10 leviathan, hi. Jun 30 18:45:18 mrmoku, j/k, no idea what it does :) Jun 30 18:45:26 mrmoku, I think it's just another not so cool example. Jun 30 18:45:30 but I might be wrong Jun 30 18:45:36 as I did notice new functions Jun 30 18:45:40 mrmoku, TAsn: I've seen that we have a folder called eclips in folder BROKEN Jun 30 18:45:41 on svn Jun 30 18:45:50 whats the meaning of that? Jun 30 18:46:13 leviathan_: svn? what svn? Jun 30 18:47:10 leviathan, talking about e? Jun 30 18:47:33 ahh... Jun 30 18:49:22 I was looking for a presentation part, where I show, how simple you can create new apps for SHR Jun 30 18:49:26 :-) Jun 30 18:53:59 leviathan, ask in #e :P Jun 30 18:54:03 leviathan, as for shr apps Jun 30 18:54:12 I think showing using python is best Jun 30 18:54:16 but if you want c Jun 30 18:54:26 show using libphone-ui Jun 30 18:54:37 for utility functions (like calling and registering to signals) Jun 30 18:54:47 libfso-glib for fso calls Jun 30 18:54:54 and elementary/edje for UI :P Jun 30 18:55:13 but mostly, I think, dbus for showing shr UI parts :P Jun 30 18:55:30 mrmoku, with should really think about generating Jun 30 18:55:34 libshr-glib Jun 30 18:55:38 like we do for fso Jun 30 18:56:05 hmm, ok Jun 30 18:56:32 mrmoku, what do you think about libshr-glib? :P Jun 30 18:58:04 TAsn: libshr-glib? Jun 30 18:58:15 mrmoku, yeah :P Jun 30 18:58:22 We should really write it already. Jun 30 18:58:53 ahhh... now :P Jun 30 18:58:53 as it's a major part of our API (and therefore should be in the SDK) Jun 30 19:00:30 hmm Jun 30 19:00:36 libshr-glib Jun 30 19:00:57 something with elementary would be cool :-) Jun 30 19:01:35 elementary already has c API Jun 30 19:01:38 we don't. Jun 30 19:02:13 (At least not a convenient one) Jun 30 20:14:07 TAsn: efl built ok, trying on new Jun 30 20:14:18 * JaMa finally returned from work :/ Jun 30 20:19:46 hmm Jun 30 20:20:09 what would be nice Jun 30 20:20:18 is a simple application creation tool Jun 30 20:20:26 like they have in android ^^ Jun 30 20:20:39 but... much more nicer will be to show Jun 30 20:20:49 how already available apps for desktop Jun 30 20:20:58 also work on SHR on smartphones Jun 30 20:21:46 imho a toolchain for SHR, including all relevant libs (and ideally vala too) would be nice Jun 30 20:22:11 but the general mood in oe and around here seems to be to frown to toolchains Jun 30 20:24:18 zub: that's mostly because current state of sdk/nativesdk recipes in OE.. Jun 30 20:24:43 zub: would be great to generate and update SDK image with one bitbake call Jun 30 20:24:58 isnt there an OE plugin for eclipse? Jun 30 20:25:04 or something like this Jun 30 20:25:05 uhmm Jun 30 20:25:07 bitbake Jun 30 20:25:08 it is Jun 30 20:25:30 then I should try what it is, and what the functionality is Jun 30 20:25:38 so that I can present it ^_^ Jun 30 20:27:10 JaMa: what do you mean by "generate and update SDK image with one bitbake call"? Jun 30 20:27:41 I've been building some OE toolchaing for my job use. It was painful, but eventually I managet to make it produce what I needed. Jun 30 20:31:40 zub: something like this http://dominion.thruhere.net/koen/cms/simplified-development-setup-for-ngstr-m Jun 30 20:31:59 * ThibG disables swap Jun 30 20:32:04 JaMa: (from earlier) applied and pushed Jun 30 20:32:31 Weiss: thx Jun 30 20:33:14 zub: and option "full SDK for generated filesystem" is what I would like to provide for shr images Jun 30 20:33:27 zub: and it should be regenerated easily (ie after EFL bump) Jun 30 20:35:58 hm, I didn't get further than adding the extra libs into the toolchain recipe (or into the appropriate task in fact) Jun 30 20:36:03 but I've seen this in narcissus Jun 30 20:38:28 zub: It could be much better with BBCLASSEXTEND nativesdk used in more packages.. but nobody likes "janitor tasks" as announced call-for-help on OE ML.. :/ Jun 30 20:42:56 JaMa: what was the subject of that mail? Jun 30 20:43:46 yeah, I'm looking through the ML in search of that too Jun 30 20:45:01 http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg03828.html Jun 30 20:48:01 JaMa: thanks Jun 30 20:53:43 JaMa: my problem with those kind of things is lacking documentation of what to do :/ Jun 30 20:55:08 ie conversion to newstaging is described in that thread and it's really not that hard (converting all xorg recipes was about 3-4 days, build time included) Jun 30 20:55:38 ThibG: fyi gta02 can work without a battery when GSM is off. Jun 30 20:55:55 and with clever script monkeys, you don't need so much manual work for "sane" recipes and when everything goes ok Jun 30 20:56:05 ThibG: i swapped my sim several times this way. Jun 30 20:56:14 JaMa: but the BBCLASSEXTEND for nativesdk ? I would be more interested in that one... Jun 30 20:56:25 hm ok Jun 30 20:56:31 * mrmoku has the SHR SDK as too long forgotten trac entry :P Jun 30 20:56:43 mrmoku: there was a thread about it about 1-2 weeks ago Jun 30 20:57:06 http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg05056.html Jun 30 20:57:33 and today again as follow-up to http://patchwork.openembedded.org/patch/2279/ Jun 30 20:57:34 Re: [oe] BBCLASSEXTEND sdk vs. nativesdk Jun 30 20:57:36 this one? Jun 30 20:57:57 yup Jun 30 21:00:25 TAsn: efl bump building on buildhost.. don't expect it sooner then tomorrow morning Jun 30 21:01:53 JaMa: as soon as my build finished and I tested it I will push a patch for 2.6.32... to make headset jack work the correct way :) Jun 30 21:02:21 ok Jun 30 21:02:24 afk Jun 30 21:04:11 mrmoku: btw I've pushed "our" kernel patches to gitorious, if you want to push it there too, otherwise I'll apply&push it from OE repo after you add it Jun 30 21:04:51 JaMa: guess I won't be doing many kernel patches :P Jun 30 21:05:03 JaMa: maybe you can just take it from OE :) Jun 30 21:18:36 PaulFertser, indeed, it seems a bit abusive to call µSD "non-removable" in the FR Jun 30 21:20:34 ThibG: and god itself told to swap sd while in suspend. Jun 30 21:21:41 really? The FSM did? Jun 30 21:21:52 ThibG: i mean, it's highly possible :) Jun 30 21:22:27 JaMa, that's good enough :P Jun 30 21:22:45 great :) Jun 30 21:23:41 gena2x, got to try something then Jun 30 21:28:25 Hm, you need to supply some power through USB when swapping the µSD, no? Jun 30 21:32:43 ThibG: joking... Jun 30 21:34:08 ThibG: but if data is not syncing properly, making sd non-removable sounds like workaround. Jun 30 21:34:49 it is a workaround and it efficiently works around it Jun 30 21:35:07 [ 234.895000] mmc0: error -110 during resume (card was removed?) Jun 30 21:35:16 what I get removing a card when suspending Jun 30 21:35:21 will try switching cards Jun 30 21:35:47 JaMa: works...OEV is 3.1 ... I make it 3.2? or 4.1? Jun 30 21:37:36 guess 3.2 is fine for an added patch Jun 30 21:38:29 ThibG: but i think where is definetly more important issues which require your attention :) Jun 30 21:38:57 JaMa: btw. if you commit it to gitorious... patch is not mine... it's from lars - I just adjusted it to 2.6.32 Jun 30 21:39:07 like? Jun 30 21:39:14 ThibG: no bugs?? Jun 30 21:39:25 ThibG: how about curlim? Jun 30 21:39:45 ThibG: my sound card is also not working from example :) Jun 30 21:39:48 the 1000mA thing? Jun 30 21:39:55 ThibG: yeah. Jun 30 21:40:07 wait wait wait, I'm on glamo for now :) Jun 30 21:40:15 ThibG: in .34 seem it really restores somehow. Jun 30 21:40:40 ThibG: but i am worriing always then i see 1000ma while on usb. Jun 30 21:41:05 "it really restores"? Jun 30 21:41:06 ThibG: ok, sorry that interrupted you :) Jun 30 21:41:10 alexxy: pushed the fix for the headset... works fine... and will hit you with next kernel build :) Jun 30 21:41:20 ThibG: in .29 i had to replug USB. Jun 30 21:41:27 oh Jun 30 21:41:29 ThibG: in .34 my ssh works. Jun 30 21:42:12 well, you know, "have glamo upstream" was a short-term goal in my plan, initially ^^' Jun 30 21:42:30 ThibG: nice plan. Jun 30 21:42:33 gnight all Jun 30 21:42:45 ThibG: k, going back to my DMA idea. Jun 30 21:42:47 and already 3 weeks on this... Jun 30 21:43:06 ThibG: oh, om people worked on it for 3 years... Jun 30 21:43:14 hehe Jun 30 21:43:37 ThibG: and somehow now it is still not in good shape. Jun 30 21:43:49 glamo? Jun 30 21:43:52 ThibG: and require your work :) Jun 30 21:44:54 s/and/as/ Jun 30 21:44:55 gena2x meant: ThibG: as require your work :) Jun 30 21:45:31 * gena2x really back to DMA... Jun 30 21:48:11 PaulFertser, if you don't agree with my "So, we can safely assume that the µSD is not removable.", please comment on the patch :) Jun 30 21:51:02 anyway, good night Jun 30 23:09:12 Howdy, I'm trying to connect to my hidden wpa network with iliwi but it doesn't seem possible to enter my SSID. any hint?, pelase **** ENDING LOGGING AT Thu Jul 01 02:59:57 2010